We present the designs of probes for making critical current density (Jc) measurements on anisotropic high-temperature superconducting tapes as a function of field, field orientation, temperature and strain in our 40 mm bore, split-pair 15 T horizontal magnet. Emphasis is placed on the design of three components: the vapour-cooled current leads, the variable temperature enclosure, and the springboard-shaped bending beam sample holder. The vapour-cooled brass critical-current leads used superconducting tapes and in operation ran hot with a duty cycle (D) of ~0.2. This work provides formulae for optimising cryogenic consumption and calculating cryogenic boil-off, associated with current leads used to make J(c) measurements, made by uniformly ramping the current up to a maximum current (I(max)) and then reducing the current very quickly to zero. They include consideration of the effects of duty cycle, static helium boil-off from the magnet and Dewar (b'), and the maximum safe temperature for the critical-current leads (T(max)). Our optimized critical-current leads have a boil-off that is about 30% less than leads optimized for magnet operation at the same maximum current. Numerical calculations show that the optimum cross-sectional area (A) for each current lead can be parameterized by LI(max)/A = [1.46D(-0.18)L(0.4)(T(max) - 300)(0.25D(-0.09)) + 750(b'/I(max))D(10(-3)I(max)-2.87b') × 10⁶ A m⁻¹ where L is the current lead's length and the current lead is operated in liquid helium. An optimum A of 132 mm(2) is obtained when I(max) = 1000 A, T(max) = 400 K, D = 0.2, b' = 0.3 l h(-1) and L = 1.0 m. The optimized helium consumption was found to be 0.7 l h(-1). When the static boil-off is small, optimized leads have a boil-off that can be roughly parameterized by: b/I(max) ≈ (1.35 × 10(-3))D(0.41) l h(‑1) A(-1). A split-current-lead design is employed to minimize the rotation of the probes during the high current measurements in our high-field horizontal magnet. The variable-temperature system is based on the use of an inverted insulating cup that operates above 4.2 K in liquid helium and above 77.4 K in liquid nitrogen, with a stability of ±80 mK to ±150 mK. Uniaxial strains of -1.4% to 1.0% can be applied to the sample, with a total uncertainty of better than ±0.02%, using a modified bending beam apparatus which includes a copper beryllium springboard-shaped sample holder.