OBJECTIVE In last few years we observe, among diabetic pregnant women, much better metabolic control. At the same time the duration of pregnancy is extended, while number of premature birth is lower and consequently there are less prematurity. METHODS To establish relation between calendar duration of pregnancy and metabolic control among pregnant diabetic women. METHODS The group consisted of 489 diabetic pregnant women. The diabetic control was evaluated by calculating average daily glycemia, maximal average daily glycemia in a trimester, Schlichtkrul's factor, J indicator and average fast glucose level. We have analyzed connection between metabolic control of diabetes and calendar duration of pregnancy in consecutive years of investigation and also within diabetes class of White and Hare. RESULTS The diabetes control among women giving birth within the biological norm was satisfactory. We have shown strong connection between frequency of premature birth and bad diabetic control within all analyzed factors during second and third trimester of pregnancy. The longest average duration of pregnancy in women with diabetes within the White classes, was 40 weeks in G/A class and the shortest 36 weeks in classes R, F, RF. Within the Hare classes the longest duration of pregnancy is 39.1 weeks in GDM class and the shortest 37.9 weeks is in DM+ class. The premature birth frequency was highest in years 1987 and 1988 when the metabolic control of diabetes was not satisfactory and during following years it was between 7 and 19%. In the White classes the lowest frequency of premature birth was in group G/A: 6.6% and the highest in classes R, F, RF: 57.1%. Within other classes that parameter stays at the level 14-19% and in Hare's groups: 10.7% in GDM, 12.3% in DM and 28.6% in DM+. We observe the best metabolic control in class G/A and GDM and the poorest in the R, F, RF and DM+ classes. There is a proven strong correlation between diabetic control and duration of pregnancy in White and Hare groups, until the second trimester in classes G/B and GDM and correlation reaching the third trimester in classes R, F, RF and DM+. Within classes B and D this correlation was not so strong. CONCLUSIONS There is a strong, proven statistical correlation (< 0.01) of all analyzed factors, which reaches second trimester between the metabolic control of diabetes and premature birth. Within the diabetes classes GDM, G/B, R, F, RF and also DM+, there is proven negative statistical correlation between the metabolic control of diabetes and duration of pregnancy.