We thank Dr Burdett for his interest in and comments on our recent editorial [1]. We agree that 0.9% saline is not physiological and can produce a hyperchloraemic acidosis; for that reason, in our practice, we tend to use Hartmann’s for most indications. Rather than a defence of saline, in asking the question ‘should we be more balanced?’ we simply aimed to highlight the very real fact that, despite best intentions, chloride may not be the worst anion to include in solutions and there are quite possibly detrimental effects of other anions about which we know very little at present. Rightly or wrongly, we have used lactate for a long time and have a pretty good idea of its effects, good and bad, even in the absence of rigorous evaluation. We are less reassured by acetate, for instance. It is incorrect to state that a regulatory body would reject 0.9% saline if it was brought to the market today: the whole point is that manufacturers seem able to put almost anything in a bag of fluid with minimal evaluation and bring it to market. This is in stark contrast to the at times unreasonable scrutiny under which a drug finds itself. We feel, however, a change in emphasis may have occurred; at no point are we suggesting 0.9% saline is a superior fluid or that we should use it on the basis of cost differential. Rather, despite everything, we have very little evidence on which to base our practice. An increasing body of clinicians, including Dr Burdett, have interpreted the available evidence that 0.9% saline is ‘inferior’. The cited in-vitro canine study investigated the effects of hypertonic solutions, all of which raised plasma osmolality by 30–45 mOsmol [2]. It appears this non-0.9% saline infusion may reduce renal blood flow and glomerular filtration; what relevance this has to humans receiving 0.9% saline is unclear and it is equally plausible that this could protect an ischaemic kidney by reducing its oxygen consumption. This is of course complete conjecture and worthy of further study. Intriguingly, relative to 0.9% saline, hypertonic saline following head injury is not associated with renal failure [3], while following burns relative to lactated Ringer’s it is; further clues, nothing definitive [4]. Was it the balanced anion (lactate) or the absence of hydroxyethyl starch that avoided renal failure in VISEP [5]? We have little more idea than Dr Burdett but would humbly suggest that claiming we know the answer (prematurely) risks perpetuating potentially bad fluid prescribing. It is also of note that hypertonic acetate has confirmed its place as a potent vasodilating agent [2], again questioning its presence in a ‘balanced’ resuscitation fluid. Our point was to not accept ‘balanced’ solutions until they prove themselves; 0.9% saline is far from perfect but it persistently escapes being strongly linked to hard clinical outcomes – we don’t accept the statement that ‘balanced’ alternatives are benign. Healey’s paper [6] is another animal model and worthy of a more complete discussion: in ‘moderate’ haemorrhage (a realistic 36% blood volume loss) saline and Ringer’s lactate were identical; in ‘massive’ haemorrhage (the eyebrow raising 218% blood volume) saline was indeed more acidifying and associated with 100% mortality. The relationship to typical human practice, for us, remains obscure. Williams et al.'s paper [7] is also puzzling. One of Hartmann's solution's recognised risks, potentially shared by many metabolically active ‘balanced solutions’, is a fall in plasma osmolality. Their conclusion ‘Large volumes of lactated Ringer’s solution administered to healthy humans produced small transient changes in serum osmolality. Large volumes of sodium chloride did not change osmolality but resulted in lower pH’ is what one would expect; take your pick, neither the pH or osmolality seemed to be correlated with an endpoint. None of these references from Dr Burdett confirm the alleged harms of 0.9% saline, but rather contribute to a general mumbling of ‘dangerous saline’ that results in clinicians abandoning a proven performer, albeit with predictable problems, for anything else a manufacturer can stick in a bag of fluid to balance it up. Finally, it was suggested that our consideration of dialysis fluid was unhelpful. We apologise if this is the case, but our acceptance of almost any anion other than chloride is where we let the manufacturers off the hook. If you want a ‘balanced’ fluid then how about an isotonic one that you can run into patients at volumes approaching 100 l over 24 h, balanced by bicarbonate and not some convenient (to the manufacturer) anion? Why are we not hanging 1-l bags of renal replacement fluid in theatres? I received correspondence from a Japanese group who use occular irrigation fluid intravenously; this kind of bicarbonate based ‘balanced Ringer’s’ appeals to us. Demand more of fluids and the fluid manufacturers and certainly don’t accept the mantra that saline is so bad that anything else is better; it may not be, and we would suggest that 0.9% saline has actually served us pretty well. Whatever they may or may not prove to be, the dangers of hyperchloraemia remain very elusive to define.