Oral appliances (OAs) of various types have shown variable success in the treatment of mild-to-moderate obstructive sleep apnoea (OSA). In an OSA sample, this study evaluated the efficacy of a diagnostic trial OA (myTAP™); the efficacy of a definitive custom-fitted mandibular advancement device (MAD) (SomnoDent Flex™); and whether a trial device can be used to distinguish treatment responder from nonresponder patients. Patients underwent overnight home sleep recordings prior to and after fitting of these appliances in order to objectively assess their sleep quality in terms of polysomnographic (PSG) respiratory measures: apnoea-hypopnoea index (AHI), oxygen desaturation index (ODI), and minimum oxygen saturation (LowSpO2). 40 patients with symptomatic OSAS were enrolled, 33 males and 7 females, with a mean age of 55.6 ± 12.73 years and an initial (T0) AHI of 26.51 ± 14.79. Trial devices were used in 16 patients (AHI: 29.9 ± 19.97, ODI: 21.06 ± 16.05, and LowSpO2: 82 ± 10.22 at T0) and definitive MADs in 28 (AHI: 23.90 ± 9.19, ODI: 16.27 ± 11.34, and LowSpO2: 82.87 ± 6.04 at T0). Statistically significant decreases in AHI (9.59 ± 8.94, p < 0.0023 ) and ODI (8.20 ± 9.67, p < 0.0129 ) were observed after treatment with the trial device. Only 8 of the patients in the trial device group went on to use the definitive device. Treatment with the definitive MAD produced statistically significant decreases in AHI (11.46 ± 9.65, p < 0.0001 ) and ODI (9.10 ± 8.47, p < 0.0016 ) and a significant improvement in LowSpO2 (85.09 ± 6.86, p < 0.0004 ). Thus, both types of device proved effective in improving the PSG parameters. This study showed that introducing an easy-to-make and low-cost trial device into the therapeutic pathway of OSAS patients can circumvent the problem of individual responses to treatment by allowing effective classification of patients: in short, it allows a first distinction to be drawn between responders and nonresponders to treatment.