Contribution: This article presents the development and validation of a novel instrument, the self-efficacy for engineering design (SEED) instrument, using a sample of 257 undergraduate students who studied engineering design (ED) in their mechanical engineering program of study. The SEED instrument consists of 35 self-report survey items distributed across nine subsections representing phases of ED. Background: ED helps foster creative thinking and promotes the application of technical knowledge to real-world problems. Self-efficacy is frequently used to evaluate educational outcomes. No adequate instruments exist to measure SEED. Research Questions: (1) Do the proposed items adequately represent and measure the SEED construct, based on experts' judgments, and thus providing evidence for content validity? (2) Do the proposed items coherently cluster into categories that meaningfully represent the SEED construct, thus providing evidence for internal structure validity? and (3) Do the proposed items pertaining to each subsection of the SEED instrument have strong and consistent relations among each other, thus providing evidence for internal consistency reliability? Methodology: This study used established theoretical frameworks in ED and design thinking to develop the SEED items. Evidence for content validity came from agreements of three experts, and for internal consistency via confirmatory factor analysis (CFA) and McDonald's [omega] indices. Findings: The proposed items for the SEED instrument can be (1) used to adequately represent and measure the SEED construct, based on experts' judgments (content validity) and (2) grouped into categories that represent the SEED construct (construct validity). The proposed items pertaining to each category of the SEED instrument have strong and consistent relations among each other (internal consistency reliability).