Numerous studies have shown that making two bars parallel to each other in the haptic domain results in (often) large and systematic errors. This is most likely due to the biasing influence of the egocentric reference frame. Even presenting participants with either haptic or visual information about parallelity or direct error feedback did not result in veridical performance. The present study was set up to assess to what extent haptic performance could be improved by providing combined visual and haptic practice. Thirty-two participants (sixteen females and males) used their dominant hand to make a test bar parallel to a reference bar located at the side of the non-dominant hand. Haptic sessions (in which participants were blindfolded and had to perceive parallelity using their hands) were alternated with visual sessions (in which they could physically see both bars and could also use their eyes to perceive parallelity on the test bar without feeling the reference bar) over a series of eleven sessions. Results showed that performance in the haptic condition significantly improved as an effect of visual practice. This effect was similar in both genders. While gender differences were significant in the haptic condition, with male participants outperforming female participants, this was not the case in the visual condition. However, veridical performance was not obtained in the haptic condition for each gender and deviations were significantly larger than in the visual condition, replicating earlier findings of a rather robust influence of the egocentric reference frame in haptic parallelity matching.