Mammographic density (MD) is a well-established risk factor for breast cancer. Whether MD influences the tumor phenotype remains to be clarified. Previous studies are highly inconsistent and most lack important covariate information. This is a case-only study within a population-based case-control study. Cases were all postmenopausal women, aged 50-74 years, with incident breast cancer, diagnosed 1993-1995, and with no history of previous cancer (n = 2,720). 1,747 women with mammograms and information on tumor characteristics were included in analyses. MD was assessed using a computer-assisted thresholding technique. We used linear, logistic, and multinomial logistic regression, adjusting for possible confounders, to study density and tumor characteristics. PD was only statistically significantly associated with tumor size in our study (regression coefficient 0.031, p = 0.017). The effect of PD on tumor size was greater when mode of detection was excluded from the model (regression coefficient 0.043, p = 0.001). No other associations between PD and the tumor characteristics studied (lymph node metastasis, ER-status, PR-status, grade, and histopathological classification) were observed. In summary, PD was positively associated with tumor size in postmenopausal women. However, the relationship was at least partially confounded by mode of detection. Although there may be a true biological relationship between MD and more highly proliferative tumors, it also seems that part of this relationship is due to masking delaying diagnosis. In conclusion, PD does not seem to be differentially associated with tumor phenotype, except for tumor size, after taking mode of detection into consideration.