Hyperemesis gravidarum, excessive vomiting in pregnancy, affects approximately 0.3-3.0% of all pregnancies, but the risk is considerably higher in pregnancies following a hyperemetic pregnancy. The reported recurrence rate of hyperemesis gravidarum is wide, ranging from 15-81%, depending on study settings. Factors affecting recurrence of hyperemesis gravidarum are as yet insufficiently studied. We sought to evaluate the recurrence rate of hyperemesis gravidarum in subsequent pregnancies, to elucidate chronological patterns of recurrence of the condition, and to analyze maternal, environmental, and pregnancy-related factors associated with recurring hyperemesis gravidarum. Out of all pregnancies ending in delivery in Finland from 2004 through 2011, data of women who had at least 1 pregnancy ending in delivery following a pregnancy diagnosed with hyperemesis gravidarum were retrieved from hospital discharge register and medical birth register (1836 women, 4103 pregnancies; 1836 index pregnancies and 2267 subsequent pregnancies). The first pregnancy with hyperemesis gravidarum diagnosis was chosen as the index pregnancy, and recurrence rate was calculated by comparing the number of hyperemetic pregnancies that followed the index pregnancy to the total number of pregnancies that followed the index pregnancy. Recurrence patterns of hyperemesis gravidarum were illustrated by presenting the chronological order of the women's pregnancies beginning from the index pregnancy to the end of the follow-up period. The associations between recurring hyperemesis and age, parity, prepregnancy body mass index, smoking, marital and socioeconomic status, domicile, month of delivery, assisted reproductive technology, sex, and number of fetuses were analyzed in both the index pregnancies and in pregnancies following the index pregnancy. There were 544 pregnancies with a hyperemesis diagnosis and 1723 pregnancies without a hyperemesis diagnosis following the index pregnancies. The overall recurrence rate of hyperemesis gravidarum in pregnancies following the index pregnancy was 24%. In case of >1 subsequent pregnancy, 11% of women were diagnosed with hyperemesis in all of their pregnancies. In the index pregnancies, recurrence of hyperemesis gravidarum was more common among women with parity of 2 than parity of 1 (adjusted odds ratio, 1.33, P= .046). Overweight women (adjusted odds ratio, 0.58, P= .036) or women who smoked after the first trimester (adjusted odds ratio, 0.27, P < .001) had lower recurrence of hyperemesis. In the comparison of the subsequent pregnancies, quitting smoking in the first trimester (adjusted odds ratio, 0.32, P= .010) and smoking continued after the first trimester (adjusted odds ratio, 0.38, P= .002) were associated with lower odds of recurring hyperemesis. Female sex of the fetus was associated with higher odds of recurring hyperemesis (adjusted odds ratio, 1.29, P= .012). In the majority of pregnancies following an earlier hyperemetic pregnancy, hyperemesis gravidarum does not recur, but hyperemetic pregnancies occur in the next pregnancies with little predictability. Only few factors associated with recurring hyperemesis could be identified. Although estimating the probability of recurrence of hyperemesis gravidarum in a subsequent pregnancy based on a woman's first hyperemetic pregnancy turned out not to be feasible, it is reassuring to know that hyperemesis does not appear to become more likely with each pregnancy and that after 1 pregnancy with hyperemesis, the following pregnancy may be different.