Dunkirk marine sediments can be used as an alternative roadbed material, and a full-scale test road using this has been successfully built at Dunkirk harbour in France. In this context, the long-term deformation behaviour of Dunkirk sediments is of crucial importance in the evaluation of roadbed deformation. The standard oedometer tests were carried out to plot the e-log sigma ' and e-logt curves, which were used to determine the compression index (C-c), swelling index (C-s) and coefficient of secondary compression C-alpha during loading and unloading. The effect of lime treatment on primary and secondary compression behaviour was investigated. The soil structure influences greatly the yield consolidation stress sigma(y)', which is increased with lime content. Both sigma(y)'-e(y) (void ratio at sigma(y)') and C-c-C-s relations are analysed. It is observed that sigma(y)' is connected with e(y) by a power function and C-c/C-s varies between 0.013 and 0.1. C-alpha remains low at consolidation stress sigma ' < sigma(y)' and reaches a peak at sigma ' = (1 - 4) sigma(y)' followed by a decrease in C-alpha. Lime addition induces the peak C-alpha moving towards a larger consolidation stress of 5 sigma(y)'. Based on the C-alpha-lg(t/t(p)) curves, it is interesting to note that t = 20t(p) (end-of-primary consolidation time) is suggested as a sufficient consolidation time to study the secondary compression behaviour of marine sediments. The C-alpha/C-c ratios of various clays including Dunkirk sediments show that C-alpha/C-c changes mainly in the range of 0.02-0.05. However, the C-alpha/C-c ratio is different for lime-treated sediments (C-alpha/C-c* = 0.0164) and untreated sediments (C-alpha/C-c* = 0.0432). Two separate zones in C-alpha-C-s* plot are clearly observed for untreated and treated sediments with boundary line of C-alpha = 0.0433C(s)* - 0.0008.