Conditioning of room air in summer is important globally. Many building however do not have an energy strategy predicated on quantitative unit-operations principles. To address this, a new probabilistic model was synthesised (Chem. Eng. Sci. XXX (2017) xxx-xxx) and a quantitative on-only strategy proposed to minimize electrical energy used. Here for the first time we report an extensive commercial-scale test of this strategy over 77 contiguous days of summer (Jan. to Mar.) in a large hotel (169 rooms) in S-E Australia (latitude -37.819708, longitude 144.959936) - a hot and dry clime. Two, dimensionally-identical suites (each of two rooms), with the same fit-out and (S-E) aspect, together with identical air conditioner (8.1 kW) and 'smart' (National Meter Identification) meters to automatically transmit contiguous (24-7) electrical use (at 30 min intervals) (n = 3696). Each suite (10.164 x 9.675 m floor plan) was auto-set to a bulk interior air temperature of 22 degrees C. In one (treated) the air conditioner was operated continuously (on-only), whilst in the other (control) it was left to wide-spread industry practice of switching on whilst the room is occupied, and off, when un-occupied (on-off). The suites had standard, single-glazed pane windows or curtain wall with heat-attenuating (fabric) internal curtains. Ambient peak temperature ranged from 17.8 to 39.1 degrees C. There were 32 days with recorded rainfall. Overall occupancy rate of both suites was almost identical at 69.6 and 71.2% respectively for the treated and control suite. This coincided with a typical business period for the hotel. Based on independent electrical energy invoices, results showed the treated suite used less energy on 47 days (61%) of the experimental period. Greenhouse gases were reduced by 12%. Savings in electrical energy cost of AUD $0.75 per day (9%), averaged over the period, were found for the treated suite adopting the alternative on-only strategy. Experimental findings overall therefore support the energy strategy hypothesis that continuous on only air conditioning will use less energy in summer. The hypothesis appears generalizable and could be adapted to a range of room geometries and seasons. Results will be of immediate interest to operators and managers of large public and commercial buildings that rely on conditioning of bulk dry air in summer. Crown Copyright (C) 2018 Published by Elsevier Ltd. All rights reserved.