Home›Blog›Gantt Chart: What the Bars Show, and What They Do Not
FundamentalsBeginner
Published September 6, 2026 · By Vincent KENNEL
Gantt Chart: What the Bars Show, and What They Do Not
The bar chart is read in every programme review, and its origin story is mostly wrong. What a Gantt chart shows, what it cannot show whatever the tool draws, and why the date everyone quotes belongs to something other than a diagram.
In brief
A Gantt chart is a bar chart of schedule information: activities are listed on the vertical axis, dates run along the horizontal axis, and each activity appears as a horizontal bar placed by its start and finish dates (PMBOK Guide, 7th edition, 2021). It shows progress against plan. On its own, it does not show the links between activities.
On a contract programme, the bar chart is the artefact everyone in the room looks at and almost nobody interrogates. A bar that has slid to the right says a date has changed. It does not say why, and it does not say what the change will cost further down the schedule. That gap is a property of the representation, not a failure of the people reading it, and it is where this article starts.
What a Gantt chart is
Two axes and a bar. The activities occupy one axis, the dates the other, and a bar sits where an activity's start and finish dates place it. That is the whole of the definition PMI has carried unchanged from the PMBOK Guide, 6th edition (2017) to the PMBOK Guide, 7th edition (2021), and what follows is an unpacking of those two axes.
Bar chart and Gantt chart name the same object. That is worth settling early, because the English-language referentials do not settle it between themselves. The PMI Practice Standard for Scheduling, 3rd edition (2019) enters the object under Bar Chart and notes that it is "also known as a Gantt chart". The APM Glossary does the reverse: in the archived snapshot of 19 August 2026 it enters Gantt chart, and carries no entry for bar chart at all. Nothing rides on the choice, and a reader who goes hunting for a difference between the two terms will not find one.
How to read a Gantt chart
Take it in the order the eye takes it.
The left-hand column lists the activities, or the components of the work breakdown structure.
The time scale across the top is "a graduated marking of linear time", in hours, days, weeks, months, quarters or years, and one chart may carry more than one unit (Practice Standard for Scheduling, 3rd edition, 2019). The French professional dictionary adds something the English definitions leave out: a calendar identifies the working and non-working periods for the unit chosen (AFITEP, Dictionnaire de management de projet, 3rd edition, AFNOR, 1996). The consequence bites immediately. A ten-day activity does not cover ten cells of the scale.
A bar is a rectangular graphical object whose length is set by the activity's start and finish dates against that scale. Bars "can overlap or be displayed side by side", which is how one line comes to carry both an activity and its progress, or an activity and its baseline.
A milestone is a selected event whose date of occurrence can be identified with certainty, and which is not reached until the result is irreversibly acquired (AFITEP, 1996). Irreversibility is the criterion, and it is what separates a milestone from a date put on the chart for decoration: PMI's own definition, "a significant point or event", does not settle that. Milestones have zero duration and consume no resources (PASEG v4.0, NDIA, 2019).
A summary bar aggregates a group of related activities and displays them as one.
The baseline is the approved version of the schedule model, changeable only through formal change control, and it is what actual results are compared against. On an audited chart it appears as grey bars sitting under the activity bars (GAO, Schedule Assessment Guide, GAO-16-89G, 2015).
The status line is the vertical line marking, in the GAO's words, "the demarcation between actual work performed and remaining work". Its name is not stable. The same object is called status date, data date or time-now date, and the Practice Standard for Scheduling, 3rd edition (2019) records that the meaning of status date "varies by the brand of project management software". Two tools do not necessarily put the same thing in the same place, which is worth knowing before comparing two charts.
What it is for on a contract programme
Four uses, all of them attested in contractual referentials rather than in folklore.
Comparing plan and actual on one line. A schedule Gantt can carry baseline dates, current dates and actual progress for each activity at once (GAO-16-89G, 2015). This is not a late use of the object. As the next section shows, it is the one it was built for.
Serving as a contractual form of presentation. ECSS-M-ST-60C (2008) states that the schedule should be presented in the form of a Gantt chart. The modality matters, and it will matter again below: the schedule content and the identification of the critical path are shall, the presentation is a should.
Communicating at the right level. NASA is blunt about its own management-level Gantt summary, which "does not provide enough schedule detail for thorough schedule analysis and is not intended for use as an analysis tool" (Schedule Management Handbook, NASA/SP-2010-3403, 2011). The same split was observed in 1966: Gantt charts served the operating level, milestone charts the top, networks the middle (Holtz, An Analysis of Major Scheduling Techniques in the Defense Systems Environment, RAND RM-4697-PR, 1966). The tool is not good or bad in the abstract. It is fitted, or not, to a level of decision.
Inspecting the schedule. Running the eye down the sequence of activities is a cheap way of judging whether a calculated critical path is plausible: what should appear is a continuous waterfall of activities from the status date through to the end of the programme (GAO-16-89G, Best Practice 6).
Where the Gantt chart really comes from
The received account dates the chart to 1910. The date is real; the object it is attached to is not. 1910 is the year of the first edition of Work, Wages and Profits, a book of 194 pages carrying seven charts, as the introduction to its second edition records (Gantt, 2nd edition, The Engineering Magazine Co., 1919). It dates no diagram.
Ten dates, and the gap between them.
What Gantt set out is a principle of measurement: "all activities can be measured by the amount of time needed to perform them", and "the space representing the time unit on the chart can be made to represent the amount of activity which should have taken place in that time". The common measure, he concludes, is time (Organizing for Work, Harcourt, Brace and Howe, 1919). He never writes Gantt chart. He calls the thing "the straight line chart".
One earlier title belongs on his record. A paper of his, A Graphical Daily Balance in Manufacture, appeared in the Transactions of the American Society of Mechanical Engineers in June 1903 (DOI 10.1115/1.4060667), and is listed in a chronological bibliography of scientific management published by the New York Public Library in 1917. What we hold is that record, not the text, so nothing here rests on what the paper argues.
The problem it was built for was an ordnance problem. Wartime orders had gone from hundreds to millions of units, and charts of the usual type "were unsatisfactory because they did not sufficiently emphasize the time" (Clark, The Gantt Chart: A Working Tool of Management, The Ronald Press Company, 1922). Gantt put it as a reversal, in the summer of 1917: "We have all been wrong in scheduling on a basis of quantities. The essential element in the situation is time, and this should be the basis in laying out any program." (Alford, Henry Laurence Gantt, Leader in Industry, ASME, 1934).
Two hard dates follow, and neither is 1910. The oldest surviving chart of this kind was drawn by David B. Porter, reported to have worked with Gantt, at the Frankford Arsenal before 15 June 1917, and Alford calls it "probably the oldest Gantt Chart in existence". The first publication of one is February 1918. The name arrives later still, fixed by the title of Wallace Clark's book of 1922, published after Gantt's death on 23 November 1919.
What the object was used for, at Gantt's death, was not project planning. Alford lists four developed applications: recording the progress of work, recording machine operation, recording man performance, and showing the work ahead of a department or plant. It is a shop instrument. On the plates Alford reproduces, the bars measure quantities against time rather than activity durations, starting at the origin of the axis and reading against cumulative totals.
The bar itself is older than the name. A construction programme in bars, carrying for each line the quantity remaining and the manpower assigned, and counted in five real working days a week to absorb the weather, was published for the Langwies viaduct in 1915 (Schürch, Der Bau des Talüberganges bei Langwies, Armierter Beton, VIII. Jahrgang, Heft 10, October 1915); the programme is attributed principally to the engineer J. Müller. A fourteen-line construction schedule for a hydroelectric development was published by Hillberg in 1916 and reproduced in Haskell's How to Make and Use Graphic Charts (1919). A graphical method for showing the whole course of work was published by Adamiecki in 1909. And the founding gesture, a single time scale adopted so that "the imagination will never be imposed upon by the eye", is Priestley's, in the chart of 1769 (A Description of a New Chart of History, 4th edition, J. Johnson, 1777).
One absence is worth as much as any of those. Haskell's treatise runs to 558 pages, is devoted entirely to charts, carries a chapter on scheduling and progress, and appeared the same year as Organizing for Work. The word Gantt does not occur in it once.
How this shop instrument became today's project calendar is later, and less documented, than the story usually allows. We have no date to offer for it, and would rather say so than fill the gap.
Bars joined by arrows are not the plain Gantt chart. What most tools display is a linked bar chart, one in which the ordering relations have been made explicit. The Gantt chart in the strict sense does not show the links between activities. This is not a theoretical case: the publisher's own guide to the most widely used scheduling tool, Microsoft Project 2016 Step by Step (Microsoft Press, 2016), states that "in Gantt chart and Network Diagram views, task relationships appear as the lines connecting tasks".
Why it is read with the network, not instead of it
Start from the bar, and ask what it does not carry.
It does not carry the ordering. Holtz demonstrated this on a worked example in 1966: the degree of flexibility available in scheduling a project "cannot be readily ascertained through the use of the Gantt charts because relationships among activities in a project are not clearly revealed", and the fact that two given activities stand in a formal predecessor-successor relationship "is not revealed by the chart". The reason is structural rather than accidental. A segment placed on an axis has no graphical means of carrying a relation to another segment. Nor can the chart be repaired by adding the information: one carrying all of it "would be too cluttered to be easily read", and would still not depict the interrelationships.
What carries the ordering is the dependency network. Kelley and Walker separated the two acts when they published the critical path method: planning is "the act of stating what activities must occur in a project and in what order", scheduling "the act of producing project timetables", and the activities are represented by an arrow diagram that lets the relations among them be studied (Critical-Path Planning and Scheduling, 1959 Proceedings of the Eastern Joint Computer Conference). That paper, the birth certificate of the critical path, mentions neither Gantt nor bar charts. The opposition between the two objects came afterwards. It is not in the record.
The object that superimposes the two has a name. AFITEP (1996) gives linked bar chart in English for the diagramme de Gantt fléché: a Gantt chart in which the ordering relations have been made explicit, and which constitutes, in fact, a network. PMI has a term of its own, the time-scaled schedule network diagram: a schedule network diagram drawn so that the position and length of each activity represent its duration and, in the glossary's own words, "essentially, it is a bar chart that includes schedule network logic" (PMBOK Guide, 5th edition, 2013).
One schedule, three views.
One vocabulary note, because it is dated. Tools long displayed the network under the name PERT chart. The name is receding: the same Microsoft Project guide contains no occurrence of PERT at all, and calls the view Network Diagram. What the network is in itself, and what the four types of link between activities are, belong to the article on dependencies between tasks.
The conclusion is sourced twice, forty-two years apart. Holtz found the scheduling techniques "complementary rather than substitutes for each other". ECSS-M-ST-60C makes the same point contractually: network content and critical path identification are shall, presentation as a Gantt chart is a should. The requirement bears on the content, the recommendation only on the form.
The French professional dictionary had already written the thesis into its definition of the word schedule: a document whose form must adapt to the messages to be transmitted and the decisions to be taken, and which may bring out the logical sequence, the milestones and the resources (AFITEP, 1996). The form is a variable. It is not the object.
In short
Know which question to put to which view. How far a date has moved, and by how much, is read off the bar. What that movement does to everything downstream is read off the network. Neither view answers for the other, and the practical condition for reading either one is knowing what it does not carry.
Stop guessing. See the real impact.
Frequently asked questions
Q.Is a Gantt chart the same thing as a project schedule?
No. The schedule is the model, with its activities, durations and ordering. The Gantt chart is one presentation of it: the PMI Practice Standard for Scheduling, 3rd edition (2019) ranks the bar chart, the network diagram and the activity list alike, as examples of presentations.
Q.Who invented the Gantt chart?
Less settled than the name suggests. Gantt developed a family of shop charts built on time as the common measure, but never used that name. It comes from the title of Wallace Clark's book of 1922, published after Gantt's death, and bars on a time scale were already in print in 1915 and 1916.
Q.Gantt chart or PERT chart: what is the difference?
Two views of one schedule, not two tools to choose between. The bars carry position and duration; the ordering is carried by the dependency network, which tools long displayed under the name PERT chart. That name is receding, and the network belongs to the article on dependencies between tasks.
References
AFNOR, AFITEP - Dictionnaire de management de projet français-anglais-espagnol - 3e édition, 1996
Allan C. Haskell - How to Make and Use Graphic Charts - 1919
APM - APM Glossary - Consulted on 2026-08-19
Carl Chatfield, Timothy Johnson - Microsoft Project 2016 Step by Step - 2016
ECSS - ECSS-M-ST-60C - Space project management, Cost and schedule management - 2008
GAO - GAO-16-89G - Schedule Assessment Guide - Best Practices for Project Schedules - 2015
Henry Laurence Gantt - Work, Wages and Profits - 2nd edition, 1919
Henry Laurence Gantt - A Graphical Daily Balance in Manufacture - ASME Transactions, vol. 24, 1903
Henry Laurence Gantt - Organizing for Work - 1919
Hermann Schürch - Der Bau des Talüberganges bei Langwies an der elektrischen Bahn Chur-Arosa, Armierter Beton, Heft 10 - 1915
J. N. Holtz - An Analysis of Major Scheduling Techniques in the Defense Systems Environment - RM-4697-PR, 1966
James E. Kelley Jr., Morgan R. Walker - Critical-Path Planning and Scheduling - 1959
Joseph Priestley - A Description of a New Chart of History - 4th edition, 1777
Karol Adamiecki - Metoda wykreślna organizowania pracy zbiorowej w walcowniach, Przegląd Techniczny - 1909
Leon Pratt Alford - Henry Laurence Gantt, Leader in Industry - 1934
NASA - NASA/SP-2010-3403 - Schedule Management Handbook - 2011
