The case-control study was conducted in Belgrade (Yugoslavia) during the period 1994-1998. The objective of the study was to investigate factors related to the occurrence of multiple myeloma (MM). The study group consisted of 100 newly diagnosed MM patients and the same number of matched hospital controls. In the analysis conditional univariate and multivariate logistic regression were applied. According to multivariate analysis the following factors were significantly related to MM: smoking > or =25 cigarettes per day (Odds ratio--OR=6.7, 95% confidence interval--95% CI=1.3-34.3); having more than two brothers (OR=2.7, 95% CI=1.3-5.3), rheumatoid arthritis in personal history (OR=4.2, 95% CI=1.2-14.8), and frequent (4-7 times per week vs. lower frequency) consumption of yogurt (OR=3.1, 95% CI=1.6-6.0) and vegetables (OR=0.4, 95% CI=0.1-1.0).