Physics problem-solving assessments often prioritize accuracy over presentation. However, the presentation quality of a student’s written solution—such as its handwriting clarity and layout—may play an important role both as an indicator and facilitator of student problem-solving and achievement. In this study, we investigated the relationships between show-work presentation, conscientiousness, and student performance in a large-enrollment introductory calculus-based physics course. Using an emergent coding process, we first iteratively developed a reliable rubric for quantifying several presentation qualities of student solutions to two introductory mechanics final exam show-work questions. We found that the coding produced scores for two orthogonal factors which we label as Form, measuring layout and legibility, and Annotation, measuring student notation of their work. Statistical analyses showed that Form was a moderate-to-weak positive predictor of various course performance measures (r ~ 0.3), including show-work score, exam score, non-exam scores, and course grade. Even when controlling for ACT Math and conscientiousness scores, Form still accounted for 4-5% of the variance of these course performance scores. In contrast, Annotation was at best very weakly and inconsistently (in sign) correlated with these performance scores. Sex differences were also observed, with female student mean scores for both Form and Annotation typically about one-half standard deviation higher than male student scores. Simultaneous mediation models of the relation between sex and course outcomes included the mediators of ACT score, conscientiousness, and Form. Results indicate that Form partially mediated the relation between sex and course grade, show-work score, multiple-choice score and non-exam score. Specifically, women tended to score higher on Form, and this effect in turn led to a positive effect on scores on all of these outcomes, typically compensating for other negative effects. The (simultaneous) mediations of ACT score and conscientiousness were much weaker, consistent with being vanishingly small. It is important to note that these are correlational results only, and control-treatment designs are called for to investigate potential causal links between Form and performance. Overall, by developing and implementing a validated rubric measuring show-work presentation quality, we find evidence that presentation quality may play a relatively small but reliable role in student performance on show-work problems and is a weak but reliable predictor of most components of course performance. These results suggest that show-work presentation may be an important factor to consider in physics instruction.