We analyze the wage effects of employment breaks of women entering motherhood using a novel within-firm matching approach where mothers' wages upon return to the job are compared with those of their female colleagues within the same firm. Using an administrative German data set we investigate three different matching procedures based on information two years before birth: (1) exact matching on individual characteristics, (2) propensity score matching and (3) a combined procedure of exact and propensity score matching. Our results yield new insights into the nature of the wage penalty associated with motherhood, since we find first births to reduce women's wages by 16 to 19 percent, regardless of the matching procedure applied. Neglecting the firm identifier and matching across all firms, however, yields a wage cut of 30 percent. Furthermore, we can show that the wage loss increases with the duration of the employment break.