This article presents the results of shake-table experiments of a 2/5 scale 3-storey reinforced concrete (RC) frame building designed with non-ductile detailing and no capacity-design considerations, tested with and without column lap-splices in the upper two storeys in two experimental series comprising different input motions. The specimen consisted of two 2-span frames in parallel jointed together by transverse beams and floor slabs. It included a set of “older” detailing comprising of: (1) plain round bars, (2) 180° end hooks in beam longitudinal reinforcement, (3) no transverse reinforcement in the joints, and (4) lap splices in potential plastic hinge zones (first test series only), which, alongside a strong-beam and weak-column mechanism, create a very fragile fuse in the panel zone region of the exterior/corner beam-column joints. In the first experimental series, the structure experienced predominantly elastic responses during the first two tests, and a top-storey rocking mechanism due to column lap-splices failure during the third test. In the second experimental series, the modified specimen with welded lap-splices experienced minor damage during the first test whereas it suffered severe shear failures in the corner beam-column joints of the first storey during the second test, associated to inter-storey drift ratios and joint rotations of up to 3.8% and 0.035 rad, respectively. The large global and local demands suffered by the specimen during this last test were associated to resonant effects, as found with the Short-Time Fourier Transform (STFT) spectrogram analysis of the building response and of the input motion.