The use of rehabilitative ultrasound imaging (RUSI) to evaluate diaphragm thickness during breathing in athletes who suffer from non-specific lumbopelvic pain presents some measurement errors. The purpose of this study was to evaluate intra- and inter-sessions, intra- and inter-rater reliabilities, and concurrent validity of diaphragm thickness measurements during breathing using transcostal RUSI with a novel thoracic orthotic device that was used to fix the US probe versus those measurements obtained using manual fixation. A total of 37 athletes with non-specific lumbopelvic pain were recruited. Intra- (same examiner) and inter-rater (two examiners) and intra- (same day) and inter-session (alternate days) reliabilities were analyzed. All measurements were obtained after manual probe fixation and after positioning the thoracic orthotic device to fix the US probe in order to correctly correlate both measurement methods. Both left and right hemi-diaphragm thickness measurements were performed by transcostal RUSI at maximum inspiration, expiration, and the difference between the two parameters during relaxed breathing. Intra-class correlation coefficients (ICC), standard errors of measurement (SEM), minimum detectable changes (MCD), systematic errors, and correlations (r) were assessed. Orthotic device probe fixation showed excellent reliability (ICC = 0.852–0.996, SEM = 0.0002–0.054, and MDC = 0.002–0.072), and most measurements did not show significant systematic errors (p > 0.05). Despite manual probe fixation with a reliability ranging from good to excellent (ICC = 0.714–0.997, SEM = 0.003–0.023, and MDC = 0.008–0.064 cm), several significant systematic measurement errors (p < 0.05) were found. Most significant correlations between both orthotic device and manual probe fixation methods were moderate (r = 0.486–0.718; p < 0.05). Bland–Altman plots indicated adequate agreement between both measurement methods according to the agreement limits. The proposed novel thoracic orthotic device may allow ultrasound probe fixation to provide valid and reliable transcostal RUSI measurements of diaphragmatic thickness during relaxed breathing thus reducing some measurement errors and avoiding systematic measurement errors. It may be advisable to measure diaphragm thickness and facilitate visual biofeedback with respect to diaphragm re-education during normal breathing in athletes with non-specific lumbopelvic pain.